Utterance Verification Using Search Confusion Rate and Its N-Best Approach

  • Kim, Kyu-Hong (School of Engineering, Information and Communications University) ;
  • Kim, Hoi-Rin (School of Engineering, Information and Communications University) ;
  • Hahn, Min-Soo (School of Engineering, Information and Communications University)
  • Received : 2005.03.28
  • Published : 2005.08.31

Abstract

Recently, a variety of confidence measures for utterance verification has been studied to improve speech recognition performance by rejecting out-of-vocabulary inputs. Most of the conventional confidence measures for utterance verification are based primarily on hypothesis testing or an approximated posterior probability, and their performances depend on the robustness of an alternative hypothesis or the prior probability. We introduce a novel confidence measure called a search confusion rate (SCR), which does not require an alternative hypothesis or the approximation of posterior probability. Our confusion-based approach shows better performance in additive noise-corrupted speech as well as in clean speech.

Keywords

References

  1. IEEE Trans. Speech and Audio Proc. v.9 no.8 Speech Recognition and Utterance Verification Based on a Generalized Confidence Score Koo, M.W.;Lee, C.H.;Juang, B.H.
  2. Proc. ICASSP Word and Phone Level Acoustic Confidence Scoring Kamppari, S.;Hazen, T.
  3. IEEE Trans. Speech and Audio Proc. v.4 no.6 Vocabulary Independent Discriminative Utterance Verification for Nonkeyword Rejection in Subword Based Speech Recognition Sukkar, R.;Lee, C.H.
  4. ISCA ITRW Automatic Speech Recognition:Challenges for the Next Millenniun The AURORA Experimental Framework for the Performance Evaluations of Speech Recognition Evaluations of Speech Recognition Systems under Noisy Conditions Hirsch, H.G.;Pearce, D.
  5. ETRI J. v.26 no.3 Filtering of Filter-Bank Energies for Robust Speech Recognition Jeong, H.Y.